Inline allocation of implicit closures.

Inline allocation of implicit closures.
Consolidate generation of closure allocation stubs into one function.
Fix type name printing when printing ast.
Committed: https://code.google.com/p/dart/source/detail?r=357

[siva, 2011-10-11 21:00:46]

LGTM

http://codereview.chromium.org/8234016/diff/1/runtime/vm/object.cc
File runtime/vm/object.cc (right):

http://codereview.chromium.org/8234016/diff/1/runtime/vm/object.cc#newcode2723
runtime/vm/object.cc:2723: }
if (!IsClosureFunction()) {
  return false;
}

instead of the explicit kind() check ?

http://codereview.chromium.org/8234016/diff/1/runtime/vm/object.h
File runtime/vm/object.h (right):

http://codereview.chromium.org/8234016/diff/1/runtime/vm/object.h#newcode1149
runtime/vm/object.h:1149: bool IsImplicitClosureFunction() const;
I am wondering if it would be more readable if we
had
bool IsRegularClosureFunction() const;
bool IsStaticImplicitClosureFunction() const;
bool IsDynamicImplicitClosureFunction() const;

http://codereview.chromium.org/8234016/diff/1/runtime/vm/parser.cc
File runtime/vm/parser.cc (right):

http://codereview.chromium.org/8234016/diff/1/runtime/vm/parser.cc#newcode1423
runtime/vm/parser.cc:1423: !func.IsClosureFunction() &&
Maybe we should append to the comment as to why
func.IsClosureFunction() needs to be in this list.

I am not sure I understand the advantage you are getting by not treating all
closure functions as static the way you had before. The assertions and closure
type could have still been gotten using the parent_function.

http://codereview.chromium.org/8234016/diff/1/runtime/vm/stub_code_ia32.cc
File runtime/vm/stub_code_ia32.cc (right):

http://codereview.chromium.org/8234016/diff/1/runtime/vm/stub_code_ia32.cc#ne...
runtime/vm/stub_code_ia32.cc:1327: } else {
add a comment // Regular closure.

[regis, 2011-10-11 23:40:19]

Thanks!

http://codereview.chromium.org/8234016/diff/1/runtime/vm/object.cc
File runtime/vm/object.cc (right):

http://codereview.chromium.org/8234016/diff/1/runtime/vm/object.cc#newcode2723
runtime/vm/object.cc:2723: }
On 2011/10/11 21:00:46, asiva wrote:
> if (!IsClosureFunction()) {
>   return false;
> }
> 
> instead of the explicit kind() check ?

Done.

http://codereview.chromium.org/8234016/diff/1/runtime/vm/object.h
File runtime/vm/object.h (right):

http://codereview.chromium.org/8234016/diff/1/runtime/vm/object.h#newcode1149
runtime/vm/object.h:1149: bool IsImplicitClosureFunction() const;
On 2011/10/11 21:00:46, asiva wrote:
> I am wondering if it would be more readable if we
> had
> bool IsRegularClosureFunction() const;
> bool IsStaticImplicitClosureFunction() const;
> bool IsDynamicImplicitClosureFunction() const;

I'd like to keep IsClosureFunction() to return true for all forms, otherwise you
need to OR all forms.

I do not like "Dynamic" (because of the coming renaming of var to Dynamic), but
prefer "Instance".

I added helpers as you suggested, bu with slightly different names.

http://codereview.chromium.org/8234016/diff/1/runtime/vm/parser.cc
File runtime/vm/parser.cc (right):

http://codereview.chromium.org/8234016/diff/1/runtime/vm/parser.cc#newcode1423
runtime/vm/parser.cc:1423: !func.IsClosureFunction() &&
On 2011/10/11 21:00:46, asiva wrote:
> Maybe we should append to the comment as to why
> func.IsClosureFunction() needs to be in this list.
> 
> I am not sure I understand the advantage you are getting by not treating all
> closure functions as static the way you had before. The assertions and closure
> type could have still been gotten using the parent_function.

I added a comment.
I prefer that is_static() reflects the static scope or instance scope of the
closure. This should make instantiation of generic closure types clearer
(upcoming cl). I felt that using is_static() to indicate the presence of a
receiver as first formal parameter was not quite correct. An instance closure
has access to the captured receiver, but via the context.

While writing this, I realized that we do not need Function::IsInStaticScope()
anymore. Removed.

http://codereview.chromium.org/8234016/diff/1/runtime/vm/stub_code_ia32.cc
File runtime/vm/stub_code_ia32.cc (right):

http://codereview.chromium.org/8234016/diff/1/runtime/vm/stub_code_ia32.cc#ne...
runtime/vm/stub_code_ia32.cc:1327: } else {
On 2011/10/11 21:00:46, asiva wrote:
> add a comment // Regular closure.

Added ASSERT.